Chap 06 Marlin 2013-Rapunzal

Process Control: Designing Process
and Control Systems for Dynamic

Performance
Chapter 6. Empirical Model
Identification
Copyright Thomas Marlin 2013

The copyright holder provides a royalty-free license for use of this
material at non-profit educational institutions
CHAPTER 6: EMPIRICAL MODEL

IDENTIFICATION
When I complete this chapter, I want to be
able to do the following.
Design and implement a good experiment

Perform the graphical calculations
Perform the statistical calculations
Combine fundamental and empirical
modelling for chemical process systems
CHAPTER 6: EMPIRICAL MODEL

IDENTIFICATION
Outline of the lesson.
Experimental design for model building

Process reaction curve (graphical)
Statistical parameter estimation
Workshop
CHAPTER 6: EMPIRICAL MODELLING

We have invested a lot of effort to learn fundamental
modelling. Why are we now learning
about an empirical approach?
TRUE/FALSE QUESTIONS
false
We have all data needed to develop a fundamental

model of a complex process
false
We have the time to develop a fundamental model of a

complex process
false
Experiments are easy to perform in a chemical process
false
We need very accurate models for control engineering
EMPIRICAL MODEL BUILDING PROCEDURE

A priori knowledge
Start
Experimental Design
Not just
process
control
Plant Experimentation
Determine Model Structure
Parameter Estimation
Diagnostic Evaluation
Alternative
data
Model Verification
Complete

Start
Looks very general; it is!

However, we still need to
understand the process!
Experimental Design
T
A
Model Verification
Complete
Changing the temperature 10 K in a ethane pyrolysis

reactor is allowed.
Changing the temperature in a bio-reactor could kill
micro-organisms

Start
Experimental Design
Model Verification
Complete
Base case operating conditions

Definition of perturbation
Measures
Duration
Safely
Small effect on product quality
Small effect of profit
We will stick with linear.
What order, dead time, etc?

Start
Experimental Design
Gain, time constant, dead time ...
Does the model fit the data used

to evaluate the parameters?
Model Verification
Complete
Does the model fit a new set of

data not used in parameter
estimation.

Start
Experimental Design
Model Verification
Complete
What is our goal?

We seek models good enough for
control design, controller tuning,
and process design.
How do we know?
Well have to trust the book and
instructor for now. But, we will
check often in the future!

Process reaction curve - The simplest and most often used
method. Gives nice visual interpretation as well.
1. Start at steady state

2. Single step to input
3. Collect data until
steady state
4. Perform
calculations
Process reaction curve - Method I
45
15
S = maximum slope
35
11
25
K p /
15
/S
shown in figure
-1
-5
-5
0
10
20
time (min)
30
Data is plotted in deviation variables
40
o u tp u t va ri a b l e i n d e vi a ti o n (K )
i n p u t va ri a b l e i n d e vi a ti o n (% o p e n )
Process reaction curve - Method II
45
15
K p /
35
11
1.5 (t63% t 28% )
0.63
t 63%
25
0.28
15
-1
t28%
-5
0
10
t63%
-5
20
time (min)
30
Data is plotted in deviation variables
40
55
Lets get get out the calculator

and practice with this
experimental data.
47
43
55
39
45
0
10
20
time
30
40
output variable, degrees C
input variable, % open
51

Process reaction curve - Methods I and II
The same experiment in either method!
Method I
Method II
Developed first
Developed in 1960s
Prone to errors
because of evaluation
of maximum slope
Simple calculations
Recommended
Experimental Design
45
15
35
11
25
15
-1
-5
-5
0
10
20
time (min)
30
Model Verification
Complete
Is this a well designed

experiment?
40
Start
Process reaction curve
Experimental Design
45
15
35
11
25
15
-1
-5
-5
0
10
20
time (min)
30
40
Model Verification
Complete
Input should be close to a perfect

step; this was basis of equations. If
not, cannot use data for process
reaction curve.
Start

Process45reaction curve
Experimental Design
i n p u t va ri a b l e , % o p e n
Start
35
11
25
15
-1
-5
-5
0
10
20
time
30
Model Verification
Complete
Should we use this data?
40
o u tp u t va ri a b l e , d e g re e s C
15
Start
Experimental Design

45
15
35
11
25
15
-1
-5
-5
0
10
20
time
30
Model Verification
Complete
The output must be moved

enough. Rule of thumb:
Signal/noise > 5
40

Start
45
10
35
25
15
-2
-6
Experimental Design
-5
Model Verification
Complete
-10
0
20
40
time
60
Should we use this data?
80

Output did not
return close to the
10
initial value,
although input
returned to initial6
value

Start
45
Experimental Design
35
25
15
-2
-6
-5
Model Verification
Complete
-10
0
20
40
time
60
80
This is a good experimental design; it checks

for disturbances

Start
Plot measured vs predicted
Experimental Design

Model Verification
Complete
15
measured
35
11
25
15
predicted
-1
-5
-5
0
10
20
time
30
40
45

Statistical method
Provides much more general approach that

is not restricted to
step input
first order with dead time model
single experiment
large perturbation
attaining steady-state at end of experiment
Requires
more complex calculations

Statistical method
The basic idea is to formulate the model so that
regression can be used to evaluate the parameters.
We will do this for a first order plus dead time model,
although the method is much more general.
How do we do this for the model below?
dY (t )
Y (t ) K p X (t )
dt
Y (s) K pe
X (s) s 1

Statistical method
We have discrete measurements, so lets express the
model as a difference equation, with the next prediction
based on current and past measurements.
'
Yi 1 predicted
a Yi
'
measured
a e t /
b K p (1 e t / )
/ t
'
X i measured
min
Y
'
predicted
Yi
'
measured
Now, we can solve a standard regression problem to

minimize the sum of squares of deviation between
prediction and measurements. For a first-order model with
dead time:
Y
'
predicted
Yi
'
measured
a (Y 'i ) m b( X 'i ) m (Y 'i 1 ) m
We evaluate the minimum by setting the partial derivatives

w.r.t. parameters a and b to zero..
EMPIRICAL MODEL
BUILDING
PROCEDURE
The model parameters
are evaluated at
numerous values of the
dead time, = /t,
and the result giving
the lowest sum of error
squared is taken as the
best estimate of all
parameters.
Always compare with
data.

Predicted vs. Measured using the statistical method
45
15
35
11
25
15
Determine Model
Structure
Experimental Design
-1
-5
-5
Model Verification
0
Complete
10
20
time
30
40
Start

Statistical method
Experimental Design
Y
'
m e a su re d o u tp u t - p re d i c ti o n , d e g re e s
Start
'
Y
i
predicted
measured
Random?
1.5
0.5
-0.5
-1
Model Verification
-1.5
0
Complete
10
20
time
30
Plotted for every measurement (sample)
40

We performed a process reaction curve for the isothermal
CSTR with first order reaction. The dynamic parameters are
C A
kmol / m 3
Kp
0.50
C A0
kmol / m 3
12.4 min
Recently, we changed the

feed flow rate by -40%
and reached a new
steady-state. What are
the CA0CA dynamics
now?
F
CA0
CA
A B
rA kCA
dC' A
V
F(C' A0 C' A ) Vk C'A
dt
dC A'
C A' KC A' 0
dt
V
F
with
and K
F kV
F kV

Match the method to the application.
Feature
Input magnitude
Experiment duration
Input change shape

Signal/noise > 5
Reach steady state
Nearly perfect step
Model structure
Accuracy with
unmeasured
disturbances
Diagnostics
Calculations
First order with dead time

Poor with significant disturbance
Plot prediction vs data
simple
Statistical method
Can be much smaller
Steady state not required
Arbitrary, sufficient information
required
General linear dynamic model
Poor with significant disturbance
Plot residuals
Requires spreadsheet or other
computer program
EMPIRICAL MODEL BUILDING

How accurate are empirical models?
Linear approximations of non-linear processes
Noise and unmeasured disturbances influence data
Lack of consistency in graphical method
lack of perfect implementation of valve change
sensor errors
Lets say that each parameter has an error

20%. Is that good enough for
future applications?
CHAPTER 6: EMPIRCAL MODELLING WORKSHOP 1

We introduced an impulse to the process at t=0. Develop and apply a
graphical method to determine a dynamic model of the process.
output
0
0
10
15
20
25
30

State whether we can use a first order with dead time model
for the following process. Explain your answer.
F0 ( s )
Gvalve ( s )
.10
v( s)
m3
% open
T1 ( s )
G tank1 ( s )
F0 ( s )
3
m
1.2 K /
250 s 1
Gsensor ( s )
Gtank2 ( s )
T2 ( s ) 1.0 K / K
T1 ( s )
300 s 1
Tmeasured ( s )
T2 ( s )
1 .0 K / K
10 s 1
(Time in seconds)

We are familiar with analyzers from courses on analytical
chemistry. In an industrial application, we can extract
samples and transport them to a laboratory for
measurement.
What equipment is
required so that could
we can achieve faster
measurements for use
in feedback control?

We are performing an experiment, changing the reflux flow
and measuring the purity of the distillate. Discuss the
processes that will affect the empirical dynamic model.
Fresh
feed
flow is
constant
Pure
product
Reactor
Mostly
unreacted
feed
CHAPTER 6: EMPIRICAL MODEL IDENTIFICATION

When I complete this chapter, I want to be
able to do the following.
Design and implement a good experiment

Perform the graphical calculations
Perform the statistical calculations
Combine fundamental and empirical
modelling for chemical process systems
Lots of improvement, but we need some more study!
Read the textbook
Review the notes, especially learning goals and workshop
Try out the self-study suggestions
Naturally, well have an assignment!
CHAPTER 6: LEARNING RESOURCES
SITE PC-EDUCATION WEB

- Instrumentation Notes
- Interactive Learning Module (Chapter 6)
- Tutorials (Chapter 6)
Software Laboratory
- S_LOOP program to simulate experimental step
data, with noise if desired
Intermediate reference on statistical method

- Brosilow, C. and B. Joseph, Techniques of ModelBased Control, Prentice-Hall, Upper Saddle River,
2002 (Chapters 15 & 16).
CHAPTER 6: SUGGESTIONS FOR SELF-STUDY

1. Find a process reaction curve plotted in Chapters 1-5 in
the textbook. Fit using a graphical method.
Discuss how the parameters would change if the
experiment were repeated at a flow 1/2 the original
value.
2. Estimate the range of dynamics that we expect from
a. flow in a pipe
b. heat exchangers
c. levels in reflux drums
d. distillation composition
e. distillation pressure
3. Develop an Excel spreadsheet to estimate the
parameters in a first order dynamic model.

Chap 06 Marlin 2013-Rapunzal

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chap 06 Marlin 2013-Rapunzal

Uploaded by

Copyright:

Available Formats

Process Control: Designing Process

and Control Systems for Dynamic

Copyright Thomas Marlin 2013

CHAPTER 6: EMPIRICAL MODEL

Design and implement a good experiment

CHAPTER 6: EMPIRICAL MODEL

Experimental design for model building

CHAPTER 6: EMPIRICAL MODELLING

We have all data needed to develop a fundamental

We have the time to develop a fundamental model of a

Experiments are easy to perform in a chemical process

We need very accurate models for control engineering

EMPIRICAL MODEL BUILDING PROCEDURE

EMPIRICAL MODEL BUILDING PROCEDURE

Looks very general; it is!

Changing the temperature 10 K in a ethane pyrolysis

EMPIRICAL MODEL BUILDING PROCEDURE

Base case operating conditions

EMPIRICAL MODEL BUILDING PROCEDURE

Gain, time constant, dead time ...

Does the model fit the data used

Does the model fit a new set of

EMPIRICAL MODEL BUILDING PROCEDURE

What is our goal?

EMPIRICAL MODEL BUILDING PROCEDURE

1. Start at steady state

Process reaction curve - Method I

Data is plotted in deviation variables

EMPIRICAL MODEL BUILDING PROCEDURE

Process reaction curve - Method II

1.5 (t63% t 28% )

Data is plotted in deviation variables

EMPIRICAL MODEL BUILDING PROCEDURE

Lets get get out the calculator

output variable, degrees C

input variable, % open

EMPIRICAL MODEL BUILDING PROCEDURE

EMPIRICAL MODEL BUILDING PROCEDURE

Is this a well designed

Process reaction curve

EMPIRICAL MODEL BUILDING PROCEDURE

Input should be close to a perfect

Process reaction curve

EMPIRICAL MODEL BUILDING PROCEDURE

Should we use this data?

Process reaction curve

The output must be moved

EMPIRICAL MODEL BUILDING PROCEDURE

EMPIRICAL MODEL BUILDING PROCEDURE

Should we use this data?

EMPIRICAL MODEL BUILDING PROCEDURE

Process reaction curve

This is a good experimental design; it checks

EMPIRICAL MODEL BUILDING PROCEDURE

Plot measured vs predicted

Determine Model Structure

EMPIRICAL MODEL BUILDING PROCEDURE

Provides much more general approach that

more complex calculations

EMPIRICAL MODEL BUILDING PROCEDURE

EMPIRICAL MODEL BUILDING PROCEDURE

EMPIRICAL MODEL BUILDING PROCEDURE

Now, we can solve a standard regression problem to

a (Y 'i ) m b( X 'i ) m (Y 'i 1 ) m